Integrated omics reveal phenylpropanoid biosynthesis as a key defense pathway against Meloidogyne incognita in Cucumis metuliferus

Abstract

The southern root-knot nematode (Meloidogyne incognita, M. incognita) poses a severe threat to cucurbit production, yet no resistant germplasm exists among cultivated species. Although Cucumis metuliferus (C. metuliferus), a wild Cucurbitaceae species, exhibits resistance to M. incognita and prior transcriptomic studies have identified key responsive genes and signaling pathways, an integrated understanding of the dynamic transcriptional and metabolic reprogramming is still lacking. Here, an integrated transcriptomic and metabolomic analysis was performed on resistant (CM064) and susceptible (CM083) C. metuliferus accessions during M. incognita infection. The resistant CM064 exhibited distinct temporal shifts in defense strategy: early-stage activation of defense pathways (hormone signaling and chitin response) coupled with repression of cell wall biogenesis, and late-stage upregulation of cell wall reinforcement, redox homeostasis, and restriction of phloem transport. In contrast, the susceptible CM083 showed progressive susceptibility, characterized by early-stage modulation of cytoskeletal activity and defense suppression, followed by late-stage upregulation of amino acid biosynthesis and downregulation of serine/threonine kinase signaling. Multi-omics integration revealed a sustained activation of the phenylpropanoid biosynthesis pathway during infection. Ectopic overexpression of Cinnamate 4-hydroxylase (C4H) was associated with significant resistance to M. incognita in cucumber. Collectively, this study provides insights into the multi-layered defense mechanism of C. metuliferus against M. incognita and supports the important role of phenylpropanoid pathway activation in nematode resistance, suggesting a potential strategy for crop improvement.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08301-7.

Induction

Root-knot nematodes (Meloidogyne spp., RKNs ) are devastating soil-borne pathogens that infect a wide range of cultivated crops, resulting in substantial economic losses [33]. Among them, the southern root-knot nematode (Meloidogyne incognita) is regarded as one of the most destructive species [1, 39]. The infection process involves root invasion, feeding site establishment, and subsequent reproduction. Second-stage juveniles (J2) penetrate the roots through the tips and migrate toward the vascular cylinder. Upon reaching the vascular cylinder, RKNs induce the formation of giant cells (GCs), which serve as nutrient sinks supporting nematode development from J3 to adult stages. Mature females eventually extrude egg masses outside the roots, thereby completing their life cycle [34, 40]. Although nematicides and soil fumigants are widely used to control nematodes, these chemical pesticides pose serious environmental and food safety concerns [12, 15]. Consequently, identifying resistant cultivars and elucidating the molecular mechanisms underlying host resistance represent the most effective and sustainable strategies for nematode control.

Cucumber (Cucumis sativus L.) and other cultivated Cucurbitaceae crops are economically important yet highly susceptible to RKN infection [29]. To date, no RKN-resistance germplasm has been identified among cultivated Cucurbitaceae species [26, 48]. Several RKN-resistance genes (R genes) have been characterized in other crops such as tomato, pepper, plum, and eggplant [9, 27, 52, 56]. The Mi-1.2 gene, encoding a nucleotide-binding site and leucine-rich repeat (NBS-LRR) protein, remains the only commercially deployed R gene in tomato. However, the Mi-1.2-mediated resistance is compromised at soil temperatures above 28 °C. Subsequent studies revealed that Mi-1.2 confers resistance through the induction of reactive oxygen species (ROS) bursts that inhibit nematode development [18]. More recently, the heat-stable Mi-9 gene was identified as conferring resistance to M. incognita through rapid ROS burst and localized cell death [7]. Notably, no orthologous sequences corresponding to these nematode-resistance R genes have been detected in cucumber and other cultivated Cucurbitaceae genomes, suggesting that these species may rely on lineage-specific defense mechanisms against nematode infection.

Wild species often exhibit superior natural resistance, providing valuable resources for uncovering resistance mechanisms against pathogens, like M. incognita. Previous studies have reported that Cucumis metuliferus, a wild relative within the Cucurbitaceae family, exhibits resistance to M. incognita, primarily by inhibiting nematode development and reproduction [25, 48]. However, the commercial utilization of C. metuliferus as a rootstock for RKN control remains limited [17]. To date, the identification of resistance-associated genes from resistant germplasm and the elucidation of their associated metabolic pathways remains central to breeding strategies for resistance. Therefore, deciphering the molecular and physiological defense mechanisms of C. metuliferus could provide valuable insights for enhancing RKN resistance in cultivated cucurbit crops. Previous studies have largely relied on comparative transcriptomic analyses between resistant and susceptible germplasm. Such studies have shown that RKN infection triggers transcriptional reprogramming in host plants, affecting diverse biological processes, including the cell cycle, cell wall remodeling, primary and secondary metabolite pathways, and phytohormone biosynthesis and signaling pathways [19, 38]. Transcriptomic analyses in C. metuliferus have revealed similar defense-associated responses, particularly involving Ca²⁺ signaling, JA/SA/IAA signaling pathways, carbohydrate metabolism, and cell wall remodeling during M. incognita infection [23, 24]. A recent comparative study identified the ten most upregulated differentially expressed genes between the resistant (CM3) and susceptible (CM27) accessions of C. metuliferus. Functional validation demonstrated that silencing Hsp20/α-crystallin family genes in CM3 significantly increased susceptibility to M. incognita [49].

However, the parasitism of root-knot nematodes involves complex and dynamic interactions between the nematodes and the host plant. An analysis dependent on transcriptomics cannot fully capture the functional biochemical outputs of these interactions. Plants deploy a wide array of metabolites as crucial defense components against pathogen attacks [4, 16]. Therefore, integrated transcriptomic and metabolomic analysis has emerged as a powerful approach to uncover the coordinated gene-metabolite networks underlying plant immunity [28, 44]. In Oryza sativa, integrated transcriptome and metabolome analysis revealed an association between flavonoid pathways and resistance against Meloidogyne graminicola [54]. In Solanum lycopersicum, comprehensive metabolome and transcriptome analyses identified a crucial role for the flavonoid pathway metabolite kaempferol in tomato-nematode interactions [55]. Although transcriptomic studies have established the resistance of C. metuliferus to M. incognita and identified candidate genes, a mechanistic insight into how transcriptional reprogramming coordinates with metabolic changes in the resistance response remains elusive.

To address this gap, this study implements a temporal, integrated transcriptomic and metabolomic analysis of resistant (CM064) and susceptible (CM083) C. metuliferus during M. incognita infection. This work aims to establish a comprehensive dynamics framework of the molecular responses to M. incognita infection and to identify key genes and metabolites associated with resistance. This study is expected to provide a foundation for elucidating the resistance mechanisms in C. metuliferus and offer insights for resistance breeding and metabolic engineering in cucurbit crops.

Materials and methods

Plant materials and growth conditions

Plants were grown in controlled growth chambers at 26 °C under a 16 h light/8 h dark photoperiod. Two inbred lines of Cucumis metuliferus CM064 (PI482440) and CM083 (PI500380), obtained from the USDA North Central Regional Plant Introduction Station (Ames, Iowa, USA), were used in this study. C. sativus (6101-4) was used for transient genetic transformation experiments. Nicotiana benthamiana was employed for transient overexpression. S. lycopersicum cv. ‘Alisa Craig’ (AC), a highly susceptible cultivar to M. incognita, was used for nematode propagation. Egg masses were separated from infected tomato roots and incubated in distilled water at 28 °C to allow for hatching. The newly hatched second-stage juveniles (J2s) were then collected for subsequent experiments.

M. incognita inoculation and tissue collection

C. metuliferus seedlings at the two-leaf stage were transferred to pots containing a sterilized peat moss-vermiculite mixture (3:1, v/v). Each plant was inoculated with about 500 second-stage juveniles (J2s) of M. incognita. The inoculation dose of approximately 500 J2s per plant is consistent with prior transcriptomic studies of C. metuliferus [24]. Nematode development was assessed by Acid-fuchsin staining [3]. The number and developmental stage of nematodes were examined under a microscope. Roots from control plants and from plants inoculated with J2s were harvested at 7 days and 14 days post inoculation (dpi). Root samples were carefully washed and collected in three biological replicates for subsequent RNA sequencing and metabolomic analyses.

Transcriptomic analysis

Total RNA was extracted from the roots of CM064 and CM083 using TRIzol reagent. All data were derived from three independent biological replicates (individual plants). After confirming RNA integrity on an Agilent 2100 Bioanalyzer, mRNA was isolated from the total RNA via poly-T oligonucleotide-coupled magnetic beads. The libraries were sequenced on lllumina Novaseq platform to generate 150 bp paired-end reads. Raw sequencing data were first processed with fastp to remove adapter-contaminated reads, poly-N sequences, and low-quality reads. Following this quality control, approximately 6-8Gb clean data were obtained for each sample (Supplementary Data S12). The reference genome and gene model annotations for C. metuliferus were acquired from a previous study [25] and downloaded from the genome website (http://cucurbitgenomics.org-/v2/ftp/genome/melon/PI482460/). Index of the genome was built using Hisat2 v2.0.5 and paired-end clean reads were aligned to the reference genome using Hisat2 v2.0.5. The resulting alignment statistics were detailed in Supplementary Data S13. Differential expression analysis was performed using DESeq2, with adjusted P-value ≤ 0.05 and |log₂ (fold change) | ≥ 1 as a significant threshold. GO enrichment for the DEGs was conducted in PANTHER v18.0 under a Fisher’s exact test with Bonferroni correction (P < 0.01), implemented via the clusterProfiler toolkit (v4.8.1). Functional classification and pathway enrichment of DEGs were analyzed using the KEGG database (http://www.genome.jp/kegg/). Heatmap of DEGs was generated by TBtools software [6].

Untargeted metabolomic analysis

The untargeted metabolomic profiling was performed on the whole roots of C. metuliferus CM064 and CM083. Roots were thoroughly washed, frozen in liquid nitrogen, and ground into a fine powder. Approximately 100 mg of tissue was homogenized in prechilled 80% methanol by vigorous vortexing. After incubation on ice for 5 min, extracts were centrifuged at 15,000 × g for 20 min at 4 °C. An aliquot of the supernatant was diluted with LC-MS grade water to a final concentration of 53% methanol, followed by a second centrifugation under the same conditions. For system conditioning and quality control, a pooled quality control (QC) sample was prepared by mixing equal volumes of all samples (n = 3), and a blank sample was prepared with 53% methanol solution.

The resulting supernatant was subjected to LC-MS/MS analysis [45]. UHPLC-MS/MS analyses were conducted using a Vanquish UHPLC system (ThermoFisher, Germany) coupled with an Orbitrap Q Exactive^TMHF mass spectrometer (Thermo Fisher, Germany). Samples were processed with Hypersil Goldcolumn (100 × 2.1 mm, 1.9 μm) using 12 min linear gradient at a flow rate of 0.2 mL/min, using eluent A (0.1% FA in water) and eluent B (Methanol). The mass spectrometer was operated in positive/negative polarity mode with spray voltage of 3.5 kV capillary temperature of 320 °C.

Raw data were processed with Compound Discoverer 3.3 (Thermo Fisher Scientific) for peak alignment, feature detection, and metabolite quantification. Raw peak area was corrected with the first QC. Peak intensity was normalized to the total spectral intensity. Metabolite annotation were as follows: the normalized data were used to predict the molecular formula based on additive ions, molecular ion peaks, and fragment ions. And then peaks were matched with mzCloud (https://www.mzcloud.org/), mzVault, and MassList database to obtain accurate qualitative and relative quantitative results.

Each annotation was assigned a confidence level: Level 1 (match in MS1, MS2, and retention time), Level 2 (match in MS1 and MS2 spectra), and Level 3 (match in MS1 only), along with a spectral matching score. Statistical analyses were conducted using R (version 3.4.3) and Python (version 2.7.6) under the CentOS operating system (release 6.6). Relative quantitation values were normalized as: [Sample Raw Value] / ([Sum of Sample Values] / [Sum of QC1 Values]). Metabolites with a coefficient of variation (CV) > 30% in QC samples were removed. Differentially accumulated metabolites (DAMs) were identified using a threshold of variable importance in projection (VIP) > 1, adjusted P-value < 0.05 (Benjamini-Hochberg FDR correction), and fold change ≥ 2 or ≤ 0.5. Principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA), and KEGG pathway enrichment analysis (adjusted P < 0.05, using Cucumis melo as reference) were conducted using the metaX toolkit [46].

Vector construction and transient transformation

The coding sequences of peroxidases (PI0005224, PI0006360, PI0000705), and CmC4H (PI0007751) were amplified from cDNA derived from C. metuliferus CM064 (Table S1). The PCR products were purified and inserted into the 35 S: PCY-HA vector containing a GFP selection marker to create overexpression vectors. Three overexpression peroxidase constructs were transformed into Agrobacterium tumefaciens strain GV3101. Three overexpression peroxidase constructs were transiently expressed in leaves of four-week-old N. benthamiana via Agrobacterium-mediated infiltration. After two days, hypersensitive response (HR)-associated cell death in Agrobacterium-infiltrated N. benthamiana leaves was observed.

Hairy root transformation and nematode resistance assay

The 35S:CmC4H-HA construct, which contained a GFP selection marker, was transformed into Agrobacterium-rhizogenes strain K599 and used to generate transgenic hairy roots in the susceptible cucumber (6101-4). At the two-cotyledon stage, cucumber roots were excised, and the seedlings were vacuum-infiltrated twice with an Agrobacterium suspension (OD600 of 0.8) for 90 seconds each time. The infiltrated seedlings were planted under high humidity (65%) in a growth chamber with 16 h light/8 h dark photoperiod at 26°C. After 4 weeks, GFP fluorescence in the hairy roots was examined, and plants with GFP-positive roots were transferred to pots containing a sterilized soil mixture (peat moss: vermiculite = 3:1). Each plant was inoculated with approximately 500 second-stage juveniles (J2s) of M. incognita. Fourteen days post inoculation, root galls were observed and quantified in GFP-positive hairy roots and non-fluorescent control roots. Three biological replicates were included in the nematode inoculation assay. To normalize gall numbers for root system size, the surface area of GFP fluorescence (used as a proxy for transgenic root biomass) on the same root segments used for gall counting was quantified from post-inoculation images using ImageJ software. Relative gall density was then calculated as the total number of galls divided by the corresponding root surface area (galls per cm²).”

Results

Phenotypic and histological characterization of M. incognita resistance in C. metuliferus

Building upon previous screening of wild Cucumis species for resistance to M. incognita [26], C. metuliferus PI482440 (CM064) and PI500380 (CM083) were selected for detailed investigation. As the resistance or susceptibility of CM064 and CM083 had not been previously reported, we first conducted a comprehensive phenotypic evaluation.

Phenotypic assessment revealed a consistent resistance phenotype in CM064 during RKN infection. The penetration rates of J2 juveniles were quantified at 7 days post-inoculation. CM064 exhibited significantly fewer penetrating J2 juveniles compared to CM083 (Fig. 1A, D, E). During subsequent nematode development, the numbers of J3 and J4 juveniles were also markedly lower in CM064 than in CM083 (Fig. 1B, F, H). Consistently, the number of root galls in CM064 was significantly reduced relative to CM083 at 14dpi (Fig. 1C, G, K) (Supplementary Fig. 4A-B). At 40dpi, CM064 has developed fewer root galls than CM083 (Supplementary Fig. S4C-D, Supplementary Fig. S5A). Together, these phenotypic data establish CM064 as resistant to RKN, a phenotype associated with impaired nematode penetration and development.

Fig. 1.

Phenotypic evaluation of M. incognita infection in resistant (CM064) and susceptible (CM083) C. metuliferus accessions. A-C Acid fuchsin staining of nematode in the resistant CM064 roots at 7 days and 14 days post inoculation (dpi). A, B Scale bar, 100 μm. C Scale bar, 1 cm. D Number of penetrating second-stage juveniles (J2s) in roots at 7 dpi (n = 5). E-G Acid fuchsin staining of the susceptible CM083 roots at 7 and 14 dpi. E-F Scale bar, 100 μm. G Scale bar, 1 cm. H Proportion of third- and fourth-stage juveniles (J3 and J4) in CM064 and CM083 at 14 dpi (n = 3). I, J Safranin O and Fast Green staining of CM064 roots at 7 and 14 dpi. I Scale bar, 100 μm. J Scale bar, 200 μm. K Number of root galls in roots at 14dpi (n = 16). L, M Safranin O and Fast Green staining of CM083 roots at 7 and 14 dpi. L Scale bar, 100 μm. M Scale bar, 200 μm. Data are presented as mean ± SE (n= number of biological replicates). Asterisks indicate a significant difference compared with the susceptible control (CM083), as determined by Student’s t-test (***P < 0.001, ****P < 0.0001)

To characterize root morphological alterations during RKN infection, histological analyses using Safranin O and Fast Green staining were performed. Both resistant CM064 and susceptible CM083 developed RKN-induced giant cells (GCs). Notably, Safranin O staining, indicating lignification, was observed surrounding GCs in CM064 at 7 dpi (Fig. 1I) and persisted at 14 dpi (Fig. 1J). In contrast, susceptible CM083 showed no detectable lignification, and nematodes developed from J2 to J4 stages, accompanied by extensive collapse of root tissue architecture (Fig. 1L, M). Taken together, lignification around GCs is associated with resistance to RKN in CM064.

M. incognita infection triggers distinct transcriptional reprogramming in resistant and susceptible C. metuliferus

Upon perceiving biotic stress, plants activate complex defense and adaptation programs. To investigate how resistant CM064 and susceptible CM083 respond to M. incognita infection at the transcriptional level, RNA sequencing of root samples at 7 days and 14 days post inoculation (dpi) was performed. Comparative transcriptome analysis revealed extensive transcriptional reprogramming and differentially expressed genes (DEGs) in both genotypes. At 7 dpi, 880 and 1475 genes were up- and down-regulated in CM083, respectively, while 1032 and 1154 genes were up- and down-regulated in CM064. At 14 dpi, 1958 and 2175 genes were up- and down-regulated in CM083, and 1776 and 1401 genes were up- and down-regulated in CM064, respectively (Fig. 2A-B, Supplementary Data S1-S4).

Fig. 2.

Transcriptomic profiling of resistant (CM064) and susceptible (CM083) C. metuliferus accessions in response to M. incognita infection. A, B Venn diagrams showing the overlap of upregulated and downregulated differentially expressed genes (DEGs) between CM064 and CM083 at 7dpi and 14dpi. T: inoculated plants; C: uninoculated control. C, D Gene Ontology (GO) enrichment analysis of the identified DEGs. Data are derived from three independent biological replicates per accession, time point (7 days and 14 days post inoculation, dpi), and condition (inoculated or uninoculated control)

For subsequent Gene Ontology (GO) enrichment analysis, only the specific (non-overlapping) DEGs identified in each genotype and time point were used to highlight unique transcriptional responses to M. incognita infection. At 7 dpi, 621 up-regulated DEGs in the susceptible CM083 were enriched in terms related to the kinesin complex, plasma membrane, and microtubule and cytoskeletal motor activity, whereas 336 down-regulated DEGs were significantly associated with defense-related pathways, including antibiotic response, secondary metabolic biosynthetic, and phenylpropanoid metabolism. In contrast, the resistant CM074 displayed up-regulation of 279 DEGs enriched in hormone-mediated signaling, chitin response, oxidoreductase activity, and serine/threonine kinase activity, while 457 down-regulated DEGs were enriched in cell wall organization or biogenesis, cell wall macromolecule metabolism, hemicellulose metabolism, cell periphery, and plasma membrane (Fig. 2C-D).

At 14 dpi, the susceptible CM083 showed 958 up-regulated DEGs mainly associated with developmental processes, including heterocycle biosynthesis, aromatic compound metabolism, organic cyclic compound biosynthesis, and cellular biosynthesis processes, whereas 979 down-regulated DEGs were enriched in phosphorus metabolism, Golgi apparatus, serine/threonine kinase activity, and root epidermal cell differentiation. Conversely, the resistant CM064 displayed 849 up-regulated DEGs associated with cell growth, cellular carbohydrate biosynthesis, cell wall polysaccharide metabolism, oxidoreductase activity, cell wall organization, and signaling receptor binding, along with 708 down-regulated genes enriched in proton transmembrane transporter, vascular transport, phloem transport, transmembrane transporter activity, and transcription factors activity (Fig. 2C-D, Supplementary Data S1-S4).

These findings indicate that the resistant CM064 and susceptible CM083 employ fundamentally different molecular patterns in response to M. incognita infection. The resistant CM064 showed a multi-layered defense response, characterized by upregulation of defense signaling and the reinforcement of structural components, along with downregulation of cell wall biogenesis and nutrient transport. In contrast, the susceptible CM083 exhibits a pattern of vulnerability, characterized by early modulation of cytoskeletal activity and suppression of defense responses, followed by sustained activation of primary biosynthesis and a progressive downregulation in defense-related signaling.

M. incognita infection triggers differential metabolic reprogramming in resistant and susceptible C. metuliferus

Metabolic changes are often associated with transcriptional reprogramming through complex regulatory networks. To characterize specific metabolic responses in resistant (CM064) and susceptible (CM083) C. metuliferus during M. incognita infection, untargeted metabolomic profiling was performed. Orthogonal partial least squares-discriminant analysis (OPLS-DA) showed significant time-dependent metabolic reprogramming, with clear separation between samples at 7 and 14 dpi (Fig. 3A). A total of 1762 differentially accumulated metabolites (DAMs) were identified across all comparisons. Among them, the resistant CM064 exhibited 84 DAMs (61 upregulated and 23 downregulated) at 7 dpi, and 549 DAMs (131 upregulated and 418 downregulated) at 14 dpi. In contrast, the susceptible CM083 showed 92 DAMs (74 upregulated and 18 downregulated) at 7 dpi, and 1037 DAMs (372 upregulated and 665 downregulated) at 14 dpi (Fig. 3B, Supplementary Data S5-S8).

Fig. 3.

Dynamic metabolic reprogramming in resistant (CM064) and susceptible (CM083) C. metuliferus accessions during M. incognita infection. A Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA) score plot of metabolic profiles. T: inoculated plants; C: uninoculated control. Points represent biological replicate (n = 3); confidence ellipses indicate 95% confidence intervals. B Number of Differentially accumulated metabolites (DAMs) identified in pairwise comparisons. C-D Classification of upregulated and downregulated DAMs at 7 dpi (C) and 14 dpi (D). E KEGG pathway enrichment analysis of the identified DAMs. Data are derived from three independent biological replicates per accession, time point (7 days and 14 days post inoculation, dpi), and condition (inoculated or uninoculated control)

Subsequent classification analysis of DAMs revealed six predominant categories with significant alterations: phenylpropanoids and polyketides, organic oxygen compounds, benzenoids, organoheterocyclic compounds, lipids and lipid-like molecules, and organic acids and derivatives (Fig. 3C-D). Notably, the resistant CM064 displayed moderate metabolic adjustments at 7 dpi, followed by substantial downregulation at 14 dpi. Conversely, the susceptible CM083 exhibited progressive metabolic dysregulation, characterized by simultaneous up- and downregulation of numerous metabolites over time.

KEGG enrichment analysis further revealed distinct pathway signatures between CM083 and CM064 (Fig. 3E). At 7 dpi, the susceptible CM083 showed enrichment in vitamin B6 metabolism and lysine degradation. In contrast, CM064 exhibited significant enrichment in phosphatidylinositol signaling, inositol phosphate metabolism, and fatty acid degradation, all of which are associated with defense-related signal transduction. At 14 dpi, CM083 showed enrichment in the biosynthesis of primary metabolites (mainly amino acids) and in arginine and proline metabolism. Meanwhile, the CM064 demonstrated enrichment in pyrimidine metabolism, tryptophan metabolism, and flavonoid biosynthesis. Together, these results indicate distinct metabolic responses: the resistant CM064 is dominated by defense-related metabolism, whereas susceptible CM083 undergoes widespread metabolic reprogramming, with a metabolic profile enriched in the amino acid metabolic pathway.

Integrated transcriptomic and metabolomic analyses implicate the phenylpropanoid-lignin pathway in resistance to M. incognita

To explore potential coordination between transcriptional and metabolic networks and identify candidate defense pathways, an integrated analysis was performed using a nine-quadrant approach. This analysis was based on DEGs and DAMs identified from the comparisons between the resistant (CM064) and susceptible (CM083) accessions at 7 dpi and 14dpi (each with three biological replicates). The analysis revealed the coordinated patterns, with Quadrant 3 (Q3) exhibiting coordinated upregulation of both DEGs and DAMs, while Quadrant 7 (Q7) showed concurrent downregulation of both (Fig. 4A-B, Supplementary Data S9-S10).

Fig. 4.

Multi-omics integration reveals coordinated defense pathways in C. metuliferus during M. incognita infection. A, B Nine-quadrant analysis integrating DEGs and DAMs (CM064_T vs. CM083_T) at 7dpi (A) and 14dpi (B). Log₂ fold changes of DEGs (x-axis) and DAMs (y-axis) are plotted. Quadrants (Q1–Q9) are defined by thresholds of ± 1.0: Q1/Q9, negative correlations; Q3/Q7, positive correlations; Q5, no correlation. C, D Correlation network of co-upregulated molecules in Q3 at 7dpi (C) and 14dpi (D). Networks display significant pairwise correlations (|r| > 0.8 and P-value ≤ 0.05). E, F Co-enrichment analysis of KEGG pathways for DEGs and DAMs at 7dpi (E) and 14dpi (F). Data are derived from three independent biological replicates per accession and condition (n = 6 per time point for correlation analysis: CM064_T, n = 3; CM083_T, n = 3)

Next, the interrelationships of co-upregulated molecules in Q3 were investigated. Pearson correlation analysis was performed on these co-upregulated DEGs and DAMs for 7dpi and 14dpi independently. Importantly, each correlation matrix was calculated separately for each time point (7dpi and 14dpi), using the combined dataset from all biological replicates of CM064 and CM083 (n = 6: three for CM064 and three for CM083). This analysis revealed significant associations between defense-related genes (e.g., PI0005224, peroxidase, PI0003666, WRKY transcription factor) and metabolites (e.g., 3-trans-caffeoyltormentic acid, 2-O-acetyl-trans-coutaric acid) (Fig. 4C-D). Separately, KEGG enrichment analysis performed separately on the DEGs and DAMs datasets indicated that the phenylpropanoid biosynthesis pathway was consistently co-enriched at both 7dpi and 14dpi (Fig. 4E-F).

Further analysis focused on the phenylpropanoid biosynthesis pathway. Key genes in phenylpropanoid-derived lignin biosynthesis pathway, including Phenylalanine ammonia-lyase (PAL), cinnamate 4-hydroxylase (C4H), 4-coumarate: CoA ligase (4CL), and cinnamyl alcohol dehydrogenase (CAD), peroxidase (POD), were significantly upregulated in CM064, while downregulated in CM083 (Fig. 5A). Correspondingly, phenylpropanoid intermediates such as ferulic acid 4-O-sulfate, ferulic acid methyl ester, methyl caffeate, (E)-methyl 4-coumarate, and 2-O-acetyl-trans-coutaric acid accumulated to significantly higher levels in CM064 at both 7 and 14 dpi (Fig. 5B-C). These coordinated transcriptional and metabolic changes are associated with resistance against RKN infection in CM064.

Fig. 5.

Integrated transcriptomic and metabolomic analyses of the phenylpropanoid biosynthesis pathway during M. incognita infection. A Schematic of the phenylpropanoid biosynthesis pathway with heatmaps of associated genes expression in CM083 (left) and CM064 (right) at 7dpi and 14dpi. Key enzyme abbreviations are defined in the figure. B, C Heatmaps showing the accumulated patterns of DAMs within the phenylpropanoid pathway in CM064 and CM083 at 7dpi (B) and 14 dpi (C). Data are derived from three independent biological replicates per condition

Ectopic overexpression of CmC4H is associated with reduced gall formation in cultivated cucumber (Cucumis sativus L.)

Previous studies have shown that cinnamate 4-hydroxylase (C4H) acts as a key enzyme in the phenylpropanoid-derived lignin biosynthesis pathway [41]. A mutation in the C4H gene results in reduced lignin deposition [35] . To further investigate the role of the phenylpropanoid-lignin pathway in defense against M. incognita, an Agrobacterium rhizogenes-mediated transient transformation system was utilized. A 35 S:CmC4H-HA (PI0007751) overexpression vector, containing a GFP selection marker for transgenic hairy roots identification, was introduced into cucumber. After 4 weeks, plants with GFP-positive roots were selected and inoculated with M. incognita. For each successfully transformed composite plant, GFP-positive roots and GFP-negative roots on the same plant served as the paired overexpression and internal control group, respectively. Fourteen days post inoculation, three independent biological replicates were analyzed, and the number of root galls on transgenic and control roots from the same plant was counted and compared. A significant reduction in total root gall formation was observed in the CmC4H-overexpressing hairy roots compared to the internal control roots on the same plant (Fig. 6A-I, Supplementary Figure S1A-C). Furthermore, when normalized to the surface area of GFP-positive roots, the gall density (galls per cm²) was also significantly lower in the overexpression lines (Supplementary Figure S3A-C). This gain-of-function analysis demonstrates that ectopic expression of CmC4H in cucumber enhances defense against M. incognita.

Fig. 6.

Ectopic expression of CmC4H in cucumber enhances resistance to M. incognita. A-F Representative root phenotype showing gall formation at 14 dpi on composite cucumber plants with transgenic hairy roots overexpressing CmC4H-GFP (OE) or non-fluorescent control roots. Scale bar, 1 cm. G-I Quantification of root galls from the experiment shown in A-F. Data are presented as the mean number of galls per root system ± SE (n = 3). Asterisks indicate a significant difference between the CmC4H overexpression (OE) and control groups, as determined by Student’s t-test (*P < 0.05, **P < 0.01)

Discussion

Dynamic defense reprogramming in resistance C. metuliferus against M. incognita infection

The parasitic interaction between root-knot nematodes (RKNs) and host plants represents a complex, long-term biological process. This process demands sustained and dynamic defense responses from the host. An integrated multi-omics analysis compared resistant (CM064) and susceptible (CM083) accessions of Cucumis metuliferus, revealing distinct molecular patterns underlying their different responses to M. incognita infection. The resistant CM064 demonstrated significantly impaired nematode development and reduced root gall formation compared to the susceptible CM083 (Fig. 1A-H, K). These phenotypic patterns align with previous reports on C. metuliferus resistance to M. incognita [23, 24]. Earlier studies relied mainly on interspecific comparisons between C. metuliferus and cultivated cucumber (C. sativus). In contract, the intraspecific data from this study provide more robust evidence by minimizing potential confounding factors arising from interspecific genomic variation. Notably, Safranin O staining, indicating lignification, was specifically observed around nematode feeding sites in the resistant CM064.

Key studies have progressively uncovered a transcriptional-level molecular basis of resistance in C. metuliferus. Comparative transcriptomics between C. metuliferus and cucumber implicated early signaling responses and highlighted DEGs involved in Ca²⁺/MAPK signaling and phytohormone (JA/SA/IAA) biosynthesis at 3dpi [23]. Furthermore, a time-course transcriptomic analysis (3, 14, and 28dpi) of infected and uninfected C. metuliferus focused on the dynamic regulation of hormone-related genes and nematode-responsive transcription factors. A cross-species comparison at 14dpi further revealed that differentially expressed genes between C. metuliferus and cucumber were enriched in biological processes related to the cytoskeleton and RNA processing [24]. Additionally, a comparative transcriptome analysis between resistant C. metuliferus and susceptible cucumber provided foundational evidence linking resistance to transcriptional changes in phenylpropanoid biosynthesis pathways [49]. Recently, an intraspecific comparison between resistant (CM3) and susceptible (CM27) C. metuliferus lines further identified the top ten DEGs as novel candidate genes, including heat shock proteins (Hsp20/alpha-crystallin family) and WRKY transcription factors. Functional validation through virus-induced gene silencing of EVM0010091 (Hsp20/alpha-crystallin family) confirmed its role in M. incognita resistance [51]. Collectively, studies cited above have successfully identified responsive genes, pathways, and validated individual candidates, providing a crucial understanding based primarily on transcriptional response. However, these studies largely focused on a single omics and therefore could not capture the dynamic interplay between transcriptional reprogramming and metabolic changes during infection. Building upon these foundations, this study provides a significant advance by applying an integrated transcriptomic and metabolomic approach. This integrated multi-omics analysis reveals that defense involves the temporally coordinated reprogramming of gene expression and metabolism, and that these dynamic processes fundamentally differ between resistant and susceptible responses.

In this study, integrated transcriptional and metabolic analyses revealed that the resistant CM064 and susceptible CM083 employed fundamentally distinct molecular patterns. The resistant CM064 indicated a dynamic and multi-layered defense response. At 7dpi, CM064 prioritized defense activation over growth by upregulating genes associated with hormone-mediated signaling, chitin response, and oxidoreductase activity, while concurrently downregulating those involved in cell wall biosynthesis and remodeling (including cell wall organization or biogenesis, cell wall macromolecule metabolism, hemicellulose metabolism) (Fig. 2C-D). This pattern may constitute an unfavorable environment for nematode feeding site establishment, consistent with previous reports that transient growth suppression is associated with beneficial stress responses [2, 53]. Additionally, metabolic reprogramming in CM064 correlated with this early defense strategy, as evidenced by the rapid upregulation of defense-associated pathways, particularly phosphatidylinositol signaling and fatty acid degradation.

At 14 dpi, the resistant CM064 consolidated its defense by sustaining defense-related activities (oxidoreductase activity and signaling receptor binding) and reinforcing cell wall structures (cell growth, carbohydrate biosynthesis, cell wall organization, and polysaccharide metabolism), while downregulating nutrient allocation systems (proton transmembrane transporter, vascular transport, phloem transport, transmembrane transporter activity, and transcription factors activity) (Figs. 1I-J and 1L-M and 2C-D). Metabolic reprogramming coincided with these transcriptional patterns, with enrichment of defense-related tryptophan metabolism, flavonoid biosynthesis, and caffeine metabolism pathways. These coordinated molecular changes are consistent with the observed impairment of nematode development in CM064.

In contrast, the susceptible CM083 exhibited vulnerability traits. At 7dpi, CM083 showed upregulation of genes related to cytoskeletal activity (kinesin complexes and microtubule motor activity), and reduced expression in defense pathways (including antibiotic response, secondary metabolic biosynthetic, and phenylpropanoid metabolism) [11]. Notably, susceptible CM083 showed early enrichment in vitamin B6 metabolism and lysine degradation pathways. At 14dpi, sustained upregulation of biosynthetic processes (heterocycle/aromatic compound biosynthesis) was observed, while the downregulation of signaling components (serine/threonine kinase activity) indicated a progressive loss of host defense responses. Consistent with metabolic alternations, CM083 prioritized amino acid biosynthesis pathways, which was associated with a nutrient-enriched metabolic profile [13].

The parasitism cycle of RKN is marked by several time points, with phenotypic differences commonly observed at 3, 7, 14, and 28dpi. This study focused on 7dpi and 14dpi to investigate host responses during feeding site establishment and subsequent nematode development. The integrated analysis thus outlines the dynamic transcriptional and metabolic reprogramming associated with C. metuliferus defense nematode parasitism. However, the absence of data from the earlier penetration stage (1-3dpi) limits a definitive resolution of the transition from initial infection to established parasitism. Future studies incorporating temporal sampling from the onset of infection (1-3dpi) will be valuable to delineate these earliest defense responses and provide new insights into the initial host-nematode interaction.

Phenylpropanoid-lignin pathway activation is associated with resistance to M. incognita

The phenylpropanoid-derived lignin biosynthesis pathway plays important role in response to abiotic and biotic stresses [5, 47]. Impaired lignin biosynthesis or deposition correlates with enhanced pathogen invasion in several plant species, including maize, rice, cotton, and Arabidopsis [21, 22, 30, 50]. Only a few studies have implicated the phenylpropanoid–lignin pathway in resistance to root-knot nematodes (RKNs) [20, 51]. For instance, in resistance tomato accessions, lignin synthesis-related genes (PAL, C4H, HCT, F5H, POD) are more rapidly and strongly induced, resulting in lignin accumulation during M. incognita infection [43]. Similarly, in rice, H₂O₂ production and phenylpropanoid-derived lignin deposition have been proposed to restrict RKN infection [14]. However, an integrated understanding of how this pathway is associated with nematode resistance remains elusive.

This study suggests that activation of the phenylpropanoid-lignin biosynthesis pathway is associated with C. metuliferus resistance to M. incognita. More importantly, this integrated temporal omics approach corroborates the transcriptional induction reported in other studies and provides biochemical evidence indicating the concerted upregulation of this pathway in a resistant cucurbit. Specifically, key lignin biosynthetic genes (PAL, C4H, 4CL, CAD, POD) were significantly upregulated in the resistant CM064 but downregulated in the susceptible CM083 (Fig. 5A). Correspondingly, the accumulations of phenylpropanoid intermediates (ferulic acid 4-O-sulfate, ferulic acid methyl ester, methyl caffeate, (E)-methyl 4-coumarate, and 2-O-acetyl-trans-coutaric acid) were markedly higher in CM064 (Fig. 5B-C). Histological analysis further revealed intense lignification around giant cells in CM064, as demonstrated by Safranin O staining (Fig. 1I-J), which aligns with transcriptional and metabolic data.

Together, these multi-omics and histological observations suggest an association between the activation of the phenylpropanoid-lignin biosynthesis pathway and nematode resistance in C. metuliferus. This correlation is further supported by the finding that ectopic overexpression of CmC4H in cucumber enhanced resistance, as evidenced by a significant reduction in root gall formation upon M. incognita infection. Future loss-of-function studies will be essential to clarify the nature of this association.

Despite the observed association of the overall phenylpropanoid pathway via C4H overexpression, it was noted that peroxidases (PODs) were also significantly induced in resistant CM064. PODs are known to consume H₂O₂ to catalyze the final step of lignin polymerization [36, 37, 42]. However, other studies have suggested that PODs are also key producers of H₂O₂, a central signaling molecule in the hypersensitive response (HR) and cell death upon pathogen attack [8, 10, 31]. Recent study demonstrated that a nematode effector suppresses plant ROS burst by directly inhibiting the peroxidase activity [32]. Thus, the role of these induced PODs in CM064 raises an intriguing question regarding their specific function.

To explore the capacity of the induced PODs, the ability of PODs to induce cell death was assessed via transient overexpression in Nicotiana benthamiana. The results showed that transient overexpression of selected PODs in Nicotiana benthamiana induced visible cell death (Supplementary Figure S2A-C). It is important to note that this finding establishes only a speculative link to the defense response in CM064 roots. Currently, there is no direct evidence that such POD-triggered cell death is elicited during M. incognita infection in CM064 or specifically around nematode feeding sites. Furthermore, experimental evidence directly connecting the observed POD upregulation to enhanced lignin deposition at the cellular level in this context is also lacking. Therefore, the precise role of these induced PODs activity in resistance, whether through lignification, cell death, or both, remains for further investigation.

Conclusion

This integrated study proposes a framework associated with Cucumis metuliferus resistance to Meloidogyne incognita. By employing temporal transcriptomic and metabolomic analyses, this study suggests a multi-layered defense response against nematodes in CM064. The resistant CM064 exhibits transcriptional and metabolic reprogramming that show a prioritization of defense activation over growth during early infection (7 dpi) and a reinforcement of structural barriers with restricted nutrient allocation at the late stage (14 dpi). In contrast, the susceptible CM083 displayed metabolic reprogramming and progressive defense suppression, which were associated with successful parasitism.

The sustained induction of the phenylpropanoid-lignin pathway is strongly associated with this resistance in C. metuliferus. Results from gain-of-function experiments involving ectopic overexpression of CmC4H in cucumber are consistent with the involvement of this pathway in nematode resistance. The upregulation of peroxidases, combined with their demonstrated capacity to trigger cell death in Nicotiana benthamiana leaves, suggests a potential link between this pathway and the modulation of cell death around feeding sites. Collectively, these findings describe a shift in resistance studies from a transcriptome-centric to an integrated, multi-omics framework for C. metuliferus. Thus, this study establishes an associative framework and proposes that future studies could explore the spatial regulation of lignification during nematode infection and evaluate the potential for engineering this phenylpropanoid-lignin pathway to enhance nematode resistance in cucurbits.

Abbreviations

M. incognita

Meloidogyne incognita

C. metuliferus

Cucumis metuliferus

RKN

Root-knot nematode

dpi

Days post inoculation

DEGs

Differentially expressed genes

DAMs

Differentially accumulated metabolites

OPLS-DA

Orthogonal partial least squares-discriminant analysis

C4H

Cinnamate 4-hydroxylase

Authors’ contributions

H.R. and X.L. conceived and supervised the project. Y.W. performed most of the experiments. M. F., L.S., X.Z., T.P., and Y.S. helped with the phenotypic identification. Y.X., Y.Z., H.X., and X.M. helped with plant cultivation and management. With input from all authors, Y.W. analyzed the data and wrote the manuscript. H.R., X.L., and M.D. reviewed and edited the manuscript. All authors read and approved the content of the manuscript.

Funding

This study was supported by the National Key Research and Development Program ‘Strategic Science and Technology Innovation Cooperation’ Key Special Project (2023YFE0206900), and the 2115 Talent Development Program of China Agricultural University and 111 Project of China Agricultural University.

Data availability

The raw RNA-sequencing data described in this study have been submitted to the NCBI database, NCBI Sequence Read Archive (SRA) under BioProject accession number PRJNA1301317. The metabolomics raw data described in this study have been deposited in the NGDC OMIX database under accession code OMIX013926. Processed data are available in the Supplementary Materials.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.